Model comparison
Qwen3-Coder 480B-A35B Instruct vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 38.1 on the Noometry Index.
Last verified . 14 shared benchmarks.
Alibaba (Qwen)
StepFunSummary
- They share 14 benchmarks with published results for both. Qwen3-Coder 480B-A35B Instruct scores higher in 0 categories and Step 5 Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Step 5 Preview leads 51.8 to 35.5.
- Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $1.50 / $7.50 for Qwen3-Coder 480B-A35B Instruct.
- Step 5 Preview accepts more context: 1.02M tokens versus 262K.
- Qwen3-Coder 480B-A35B Instruct has downloadable open weights; the other is API-only.
Side by side
| Qwen3-Coder 480B-A35B Instruct | Step 5 Preview | |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 38.1 | 47.9 |
| Released | 2025-04 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 262K | 1.02M |
| Max output | 66K | 66K |
| Input $ / M tokens | $1.50 | $1 |
| Output $ / M tokens | $7.50 | $2.70 |
| Results tracked | 25 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 35.5 (#223), Step 5 Preview: 51.8 (#36)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena WebDev | 1275 | 1564 |
| LMArena Coding | 1412 | 1480 |
| SWE-bench Verified (bash only) | 55.4% | — |
| SciCode | — | 58.9% |
| GSO | 4.9% | — |
| WeirdML | 41.2% | — |
| ALE-Bench | 461.45 | — |
| AlgoTune | 1.44 | — |
Agentic & Tool Use Not comparable
Qwen3-Coder 480B-A35B Instruct: 23.9 (#123), Step 5 Preview: —
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| Terminal-Bench | 27.2% | — |
Reasoning Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 25.5 (#149), Step 5 Preview: 40.0 (#57)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Hard Prompts | 1372 | 1465 |
| Kagi LLM Benchmark | 49.5% | — |
| CritPt | — | 20.9% |
Math Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 37.6 (#150), Step 5 Preview: 46.1 (#72)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Math | 1365 | 1470 |
| ProofBench | — | 42% |
Knowledge Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 37.0 (#162), Step 5 Preview: 41.2 (#112)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1338 | 1470 |
Multimodal Not comparable
Qwen3-Coder 480B-A35B Instruct: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 47.7 (#148), Step 5 Preview: 53.8 (#54)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1346 | 1432 |
| LMArena Chinese | 1357 | 1519 |
| LMArena Russian | 1366 | 1436 |
| LMArena Spanish | 1360 | 1439 |
| LMArena French | 1398 | — |
| LMArena German | 1325 | — |
| LMArena Japanese | 1310 | — |
| LMArena Korean | 1305 | — |
Instruction Following Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 71.6 (#147), Step 5 Preview: 76.0 (#50)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1355 | 1444 |
Long Context Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 42.0 (#131), Step 5 Preview: 44.5 (#64)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1378 | 1455 |
Writing & Preference Step 5 Preview leads
Qwen3-Coder 480B-A35B Instruct: 55.3 (#147), Step 5 Preview: 62.8 (#71)
| Benchmark | Qwen3-Coder 480B-A35B Instruct | Step 5 Preview |
|---|---|---|
| LMArena Text | 1357 | 1442 |
| LMArena Creative Writing | 1333 | 1410 |
| LMArena Multi-Turn | 1365 | 1447 |
Frequently asked questions
Is Qwen3-Coder 480B-A35B Instruct better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 38.1 on the Noometry Index.
Which is cheaper, Qwen3-Coder 480B-A35B Instruct or Step 5 Preview?
Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; Qwen3-Coder 480B-A35B Instruct lists at $1.50 and $7.50.
Is Qwen3-Coder 480B-A35B Instruct or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 35.5 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 262K.
How many benchmarks do Qwen3-Coder 480B-A35B Instruct and Step 5 Preview share?
14 benchmarks have published results for both models. Qwen3-Coder 480B-A35B Instruct has 25 scored results on Noometry and Step 5 Preview has 18.